Fixes #2

Refactor code to use `EmbeddingKey` class for embedding operations.

* **Introduce `EmbeddingKey` class:**
  - Add `EmbeddingKey` class to encapsulate `text` and `model` in `src/operations.py`.
  - Update functions in `src/operations.py` to use `EmbeddingKey` instead of tuples for keys.
  - Refactor `write_embedding_to_table`, `is_key_in_table`, `list_keys_in_table`, and `get_embedding_from_table` to use `EmbeddingKey`.

* **Refactor embedding operations:**
  - Create `EmbeddingOperations` class in `src/embedding_operations.py` to encapsulate embedding-related operations.
  - Move functions `pickle_embeddings`, `duckdb_embeddings`, and `get_similarity` into `EmbeddingOperations` class.
  - Add methods for loading and saving embeddings, checking if an embedding exists, and calculating similarity.

* **Update embedding functions:**
  - Refactor `pickle_embeddings` and `duckdb_embeddings` functions in `src/embedding.py` to use `EmbeddingKey`.

* **Add error handling:**
  - Handle potential exceptions when creating embeddings and provide meaningful error messages in `src/openai_client.py`.
  - Add error handling for database operations to manage connection issues or query failures in `src/connection.py`.

* **Add configuration file:**
  - Add `config.yaml` with database name, model name, paths to files, OpenAI API key, and number of top documents.

* **Add tests:**
  - Create unit tests for `EmbeddingOperations` class in `tests/test_embedding_operations.py`.
  - Write tests for loading and saving embeddings, checking if an embedding exists, and calculating similarity.
  - Write tests to verify that the application correctly reads and applies settings from the `config.yaml` file in `tests/test_config.py`.